Burton — On the Siqyernumerari/ Rings of the Rainhow. 187 



of any of the rest, that, as near as I could judge, it was equal 

 to them all taken together." {^Philosophical Transactions, vol. xxxii., 

 p. 243.) On several occasions I have seen bows corresponding with 

 this description ; but the breadth of the colours was not always 

 the same, being sometimes very narrow, and at other times more 

 considerable. It does not often happen that so many coloured rings 

 become visible together, though in almost every rainbow I find that 

 the first additional ring of green can be observed in contact with the 

 primitive violet : the appearance of a reddish or purple ring within 

 the green is also not unusual ; and on less frequent occasions these 

 colours are repeated in the same order. In almost every instance 

 we can see that the interference rings decrease in brightness as they 

 recede from the primitive bow, and they are also narrower than the 

 rings which compose the latter. 



On the 26th of June, 1877, at 7"40 p.m., I observed a rainbow 

 with supernumeraiy rings differing in some particulars from those now 

 mentioned. In this instance there was in contact with the violet a 

 ring of red, which was followed by other rings in this order — green, red, 

 green, red, green, red. These colours were almost as bright as those 

 of the principal bow ; and although the last ring of red seemed 

 as brilliant as any of the others, there was no indication of any fainter 

 rings beyond it. They were not formed at all parts of the arc, but 

 only in positions where the sun seemed to shine strongly ; also, the 

 breadth of each ring seemed to be ecj[ual to the sun's diameter ; and 

 the colours observed were red and green — not purple and green, as 

 they usually appear. 



The additional colours on the outside of the secondary bow are very 

 rarely seen on account of their faintness ; indeed, I have not been 

 able to notice in this position more than a single green ring, and that 

 only on a few occasions. 



An important principle in connexion with the supernumerary 

 colours is, that there is a limit to the size of the drops in which 

 they are formed ; so that, although the smaller drops may produce 

 them, those of larger size are quite inelfieient to do so. This principle 

 maybe expected, from Dr. Young's theory; for it appears, from his 

 mode of calculation, that the number of returns of the rays within a 

 given distance from the caustic increases with the size of the drops : 

 consequently, the rings being more numerous in equal spaces, the 

 breadth of each must diminish as the drops grow larger ; and when a 

 certain limit is attained, the opposite colours being very close, will 

 reach the eye compound or white, just as if no interference took 

 place. That this is really the case may be shown by examining the 

 coloured rings produced by single drops of rain in the following 

 manner : — 



If we take a fine hair, or a slender fibre of flax, &c., and, holding 

 both ends of it in the fingers, immerse it in water, again withdrawing 

 it in such a manner as will cause some particles of the fluid to adhei'e to 

 it ; upon bringing one of these particles very close to the eye, and in 



